exergy analysis of glycerol steam reforming in a heat ......2021/02/01 · exergy analysis of...
TRANSCRIPT
ww.sciencedirect.com
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 7
Available online at w
ScienceDirect
journal homepage: www.elsevier .com/locate/he
Exergy analysis of glycerol steam reforming in aheat recovery reactor
Felipe Pinheiro Falc~ao Dias a, Igor Teles Fernandes a,Andr�e Valente Bueno a,*, Paulo Alexandre Costa Rocha a,Mona Lisa Moura de Oliveira b
a Universidade Federal Do Cear�a, Departamento de Engenharia Mecanica, Campus Do Pici 715, Fortaleza, CE, 60440-
554, Brazilb Universidade Estadual Do Cear�a, Centro de Ciencias e Tecnologia, Campus Itaperi, Fortaleza, CE, 60.714-903,
Brazil
h i g h l i g h t s
� Detailed exergy analysis of glycerol steam reforming.
� Syngas maximum exergy value and efficiency coincided with maximum hydrogen yield.
� Experiments indicate a maximum reforming exergy efficiency of 75.8%.
� Trade-off between irreversibilities and tar exergy losses was evidenced.
a r t i c l e i n f o
Article history:
Received 31 August 2020
Received in revised form
14 December 2020
Accepted 30 December 2020
Available online 22 January 2021
Keywords:
Exergy analysis
Heat recovery steam reforming
Hydrogen production
Glycerol reforming
Biodiesel sub-product
* Corresponding author.E-mail address: [email protected] (A.V. Buen
https://doi.org/10.1016/j.ijhydene.2020.12.2150360-3199/© 2021 Hydrogen Energy Publicati
a b s t r a c t
A detailed exergy analysis was performed for the steam reforming process of glycerol by
means of a series of experiments in a bench scale apparatus. The reforming was conducted
in a fixed bed reactor, which operated in heat recovery mode by extracting the demanded
energy from hot exhaust gases provided by a diesel engine. In order to determine the role of
the main operational parameters into the exergy efficiency of the studied process, the
experiments were carried out with glycerol feed concentrations in water ranging from 10%
up to 90% weight, whereas the outlet reactor temperature was varied from 600 �C up to
800 �C. Detailed exergy balances revealed a compromise between the exergy destruction
within the reforming reactor and liquid separator versus the exergy losses associated to the
tar and char outputs. This trade-off was favourable to the 50% and 70% glycerol feed
concentration regimes and plateaus of about 74% exergy efficiency and 24 MJ/kg dry syngas
exergy content were verified from 650 to 800 �C reactor temperatures.
© 2021 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
o).
ons LLC. Published by Elsevier Ltd. All rights reserved.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 78996
Introduction
The current fast-growing demand for renewable energy
sources is believed to be maintained as a response to the
environmental concerns raised by the widespread fossil fuels
utilisation [1]. Biofuels are an important part of this initiative
and play a key role in mobility applications, as the fossil fuel
production tends to lose dominance due to its environmental
footprint and supply insecurity [2]. Biodiesel and ethanol are
currently leading this initiative in the mobility sector. These
fuels require slight or no modifications in the physical struc-
ture of the engines and its production technologies are rela-
tively straightforward. Furthermore, the existing fuel delivery
infrastructure has been successfully applied to supply these
biofuels to the points of consumption [3].
A wide diversity of feedstock alternatives is currently
available to meet the biofuels demand when one takes bio-
diesel into account: vegetable oils [4,5], waste oil [6], animal fat
[7] and algae [8]. Despite the diversity of characteristics of the
raw materials accessible for biodiesel industrial plants, a
single reaction route, which is denominated trans-
esterification, practically dominates its large scale production
[9]. In the transesterification process a triglyceride reacts with
a short-chain alcohol in the presence of a catalyst to form
biodiesel and the sub-product glycerol. There is a general
consensus that finding an adequate destination to the huge
amounts of glycerol resulting from biodiesel production poses
a critical issue to the development of environmental friendly
and efficient biorefinery plants [10,11]. Glycerol destination is
closely related to the sustainability of biodiesel industry, a
factor that, together with deforestation and food competition
issues, is currently moulding the future of the biodiesel pro-
duction chain [12].
In order to achieve the standards required for its applica-
tion in chemical industries, the glycerol generated in biodiesel
production plants, denominated crude glycerol, demands an
expensive purification process designed to remove methanol,
catalyst and salt traces [13]. For that reason, glycerin synthe-
sized by other routes has a relatively lower cost in comparison
to the raw material obtained by recycling the crude glycerol
[14]. This scenario prevented the traditional uses of glycerol
from being capable of absorbing the overproduction coming
from biodiesel industry and posed its discharge management
as a challenge to biodiesel sustainability [15]. Therefore, it is
crucial to develop alternative applications for the crude glyc-
erol in order to limit oversupply [16], which otherwise could
threaten the environmental benefits of replacing fossil fuels
by biodiesel [17].
Thermal processes such as pyrolysis and steam reforming
have received increasing attention as means for glycerol
management in biorefineries [9,18e20]. Glycerol has high
hydrogen content [21] and can be processed to syngas at
relatively low temperatures due to the presence of oxygen on
its formulation [22]. Promising results in terms of efficiency
gains and lower emissions have been reported with the inte-
gration of glycerol reforming systems, power generating units
and biodiesel production plants [23]. According to Reza Ziyai
et al. [24], to perform the crude glycerol reforming and then
convert the produced hydrogen into electricity can increase
the economic viability of biodiesel plants. Cormos et al. [25]
evaluated the techno-economic and environmental perfor-
mances of hydrogen and power generation based on glycerol
steam reforming with and without carbon capture. These
authors demonstrated that hydrogen and power co-
generation are a promising option to further increase the
overall energy efficiency and to improve the plant flexibility.
Matson and co-workers [26] have shown that a biodiesel plant
can also become self-sufficient, from the electric power point
of view, by using syngas obtained from glycerol to operate a
dual fuel electric generator.
Beyond the opportunities of improving the biodiesel pro-
duction chain, hydrogen production via glycerol could also
take part of a smooth transition process leading to a renew-
able hydrogen economy, in contrast to the current scenario
where fossil fuels predominate as raw material. In fact, the
relevance of hydrogen energy field appears to mould the
research efforts dedicated to the syngas production from
glycerol and, in consequence, most of the attention dedicated
to the study of its thermal processing has been focused in
obtaining a high hydrogen yield. As stated in the review by
Silva et al. [27], aspects such as the optimisation of process
parameters [28e30], novel reactor designs [31e33] and the use
of advanced catalysts [34e36] have received considerable
interest.
To guarantee that the glycerol reforming process achieves
acceptable levels of energy efficiency is another relevant
operational aspect, perhaps as important as obtaining a high
hydrogen yield. Interestingly, the number of studies dedicated
to this subject is still limited and the currently available works
share the samemethodology: numerical simulations adopting
the Gibbs free energy minimization method [37e39]. In a
recent work, Pashchenko and collaborators [37] calculated the
energy efficiency of biofuels steam reforming using waste-
heat recuperation as energy source. Their results indicate a
maximum energy efficiency of 62.5% to the glycerol reform-
ing, whichwas obtained at 627 �Cwith awater to glycerol ratio
of 1:5 wt and a pressure of 1 bar. Hajjaji et al. carried a detailed
thermodynamic analysis of the hydrogen production via
thermochemical processing of glycerol by simulating both its
steam and autothermal reforming within isothermal reactors
[38,39]. They registered a maximum exergy efficiency level of
67.8% for the autothermal reforming, which was achieved at
627 �C, 1.08 water to glycerol feed ratio and 0.33 oxygen to
glycerol ratio [38]. In the steam reforming process, the authors
recommend a temperature of 827 �C and a water to glycerol
feed ratio of 1.17 that corresponds to a maximum exergy ef-
ficiency of 66.1% [39].
Previousworks demonstrate that glycerol decomposition is
an endothermic process that requires high temperature and a
considerable concentration of water to achieve adequate
levels of hydrogen production, conditions that are not
favourable to the operation of the reforming system at high
values of energy efficiency. Thus, simultaneously complying
with hydrogen yield and thermodynamic efficiency re-
quirements can be a very challenging task that would demand
an advanced diagnosis approach. Valuable information
regarding the effects of operational parameters upon the
reforming system efficiency can be obtained via exergy anal-
ysis, a well stablished thermodynamic method that is widely
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 7 8997
used for process evaluation and rationalisation [40]. Indeed,
the exergy analysis method has been successfully applied in a
number of recent studies dedicated to the reforming of coal
[41], biomass [42] and ethanol [43,44].
Motivation, novelty and research objectives
In this study, the thermodynamic efficiency of the glycerol
steam reforming was evaluated by means of a series of ex-
periments carried out in a bench-scale heat-recovery reactor.
Waste heat from the exhaust gases produced by an internal
combustion engine was used as the energy source to perform
the glycerol reforming. Besides the benefits associated to en-
ergy conservation, the use of heat recovery to proceed the
glycerol gasification also aimed at better reproducing practical
conditions where output hot gases losses and a non-
homogeneous bed temperature profile are present. Energy
and exergy balances were conducted for the complete system
and its main components with the objective of establishing
cause and effect relationships between relevant operational
parameters and the reforming thermodynamic efficiency. The
experimental procedures adopted here are also expected to
better capture the effects of tar and solid carbon (char) for-
mation in comparison to the equilibrium simulation methods
currently in use, making it possible to reach glycerol to water
feed ratios of up to 9:1 with acceptable accuracy.
Materials and methods
A series of steady state experiments was performed in a bench
scale fixed bed reforming reactor operatingwith glycerol. Each
component of the reforming system was instrumented to
provide the data required to analyse the degree of thermo-
dynamic perfection involved in the production of synthesis
gas suitable for use as a fuel. A National Instruments SCXI
system was used to perform the signal conditioning and data
acquisition tasks and the user interface was built with the
LabView software. Real time closed loop control was imple-
mented to the fuel metering and waste heat systems. A
schematic representation of the experimental apparatus
adopted in this work can be found in Fig. 1.
The reforming process configuration considered here will
be described toward the glycerol and syngas paths. A liquid
solution of glycerol (99,5% purity) into deionised water was
delivered to the reactor by means of a computer controlled
injection system composed by an automotive fuel pump, a
pressure relief valve and a solenoid operated pintle-type fuel
injector. The reactants were injected at a constant pressure
(3.5 bar(g)) and with variable flow rate, which was controlled
through the pulse widthmodulation technique applied on the
solenoid injector. Each experimental regime received a spe-
cific mass flow set point that was previously calculated in
order to maintain a constant residence time of 4s in all of the
experiments. The residence time calculations took the ther-
modynamic state of the water vapour at the reactor outlet as
reference. The reactants mass flow and density were
measured with a Siemens FC300 DN4 Coriolis flow meter that
also provided the output signal demanded for closed-loop
control of the feed injection rate.
A water-cooled head was employed to proceed with the
glycerol-water mixture injection, vaporisation and distribu-
tion across the reactor bed inlet surface. The fuel injector was
centrally mounted at the top of the head and its spray
impinged a bowl-shaped vaporisation pre-chamber operating
at ca. 400 �C. The gaseous mixture leaving the pre-chamber
was directed to the reaction bed by a series of vapour chan-
nels. Glycerol reforming occurred in a cylindrical reactor with
94 mm of internal diameter and 500 mm height. The reactor
was packed with nickel/alumina spheres of 1.5 mm diameter
with 18 2 wt% Ni content (Gunina Engineers, India). In order to
discard any influence of the catalyst deactivation effects, the
reforming reactor was filled with brand new packing material
that was in situ reduced in H2/N2 flow before each experi-
mental run. A total of 70.5 kg of packing material was
consumed through the complete set of experiments. A
filtering plate with multiple 0.4 mm orifices supported the
reaction bed.
The reforming reaction received waste-heat from com-
bustion products provided by an internal combustion engine.
The complete heat-recovery reformer had a construction
similar to a shell and tube heat exchanger, with one pass in
the inner tube corresponding to the reaction bed and 21 passes
of exhaust gases in its shell-side. Heat losses to the sur-
roundings were prevented by using a ceramic foam insulation
sleeve mounted outside the exhaust gases labyrinth. In order
to endure the high temperatures involved in the studied pro-
cess, the injection head, the heat-recovery reactor and
external tubing were all build in AISI 310 stainless steel.
Inconel k-type thermocouple probes were placed at the
evaporation chamber, reaction bed inlet, reaction bed outlet,
exhaust gases inlet and exhaust gases outlet. Pressure mea-
surements were also conducted at these points with piezor-
esistive transducers (Omega PXM409) with an accuracy of
0.11 kPa.
Reforming temperature is a fundamental operational
parameter and, accordingly, the experiments were performed
at prescribed values of reactor outlet temperature that were
maintained with a repeatability of 10 K. The DAQ and control
system adjusted this temperature by actuating in the amount
of hot exhaust gases delivered by an internal combustion
engine to the reforming reactor. The engine used for the ex-
periments was a high-speed diesel (see Table 1) and it was
operated at full throttle conditions, while the control system
actuated in the amount of brake torque that was applied by an
alternate current dynamometer. The engine exhaust gas flow
was indirectly computed with 27 g/min uncertainty from
combined measurements conducted with an Omega FMA-
900-V intake air flow meter and a Siemens FC300 fuel flow
meter. The exhaust gases temperatures at the heat recovery
circuit inlet and outlet were measured with Inconel k-type
thermocouple probes with an accuracy of 0.62 K. Further de-
tails regarding the reactor construction can be found in a
previous work of Bueno and Oliveira [22].
The high temperature wet syngas produced in the experi-
ments was diverted to a water-cooled condenser in order to
separate its gaseous fraction from the liquid and solid ones.
This device is a cylindrical separator with a fixed bed of
0.9 dm3 filled with the reactor’s packing material and sur-
rounded by a high flow cooling water jacket. The liquid
Fig. 1 e Schematic representation of the experimental setup.
Table 1 e Engine specifications.
Configuration Single Cylinder High-SpeedNaturally Aspirated
Cycle Air Cooled 4 stroke
Injection System Mechanical - Direct Injection
Bore x Stroke (mm) 78 � 62
Compression Rate 23:1
Adopted Fuel Soybean Oil Methyl Ester
Maximum Brake Power (kW) 4,25 kW @ 3694 rpm
Thermal Efficiency at
Maximum Brake Power
19.37%
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 78998
fraction mass flow and density were measured with a
Siemens FC300 DN4 Coriolis flow meter. The solid carbon
accumulated in the reactor and condenser beds was rinsed
after the end of each experimental run, separated by the paper
filter technique, dried and determined by direct weighting.
Total carbon analysiswas applied to the liquid fraction using a
HACH Spectrophotometer DR 2800 with MR direct method.
The dry syngas left the condenser at room temperature and
was directed to an exhaust device, while a small sample was
sent to a gas chromatograph (GC Varian CP 3800) in order to
determine its composition. A thermal conductivity detector
(TCD) was used for hydrogen, carbon dioxide and carbon
monoxide analysis, whereas light hydrocarbons were
computed by a flame ionisation detector (FID) with oven
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 7 8999
temperature of 350 �C and a capillary column of the CP-Sil 5 CB
type. The variability observed in the mole fractions obtained
by the gas chromatography analysis was always less than 2%.
The uncertainties associated to thermodynamic quantities
such as the syngas exergy and the exergy efficiencies were
calculated from the contributions of each measured variable
demanded for its determination, according to Ref. [45]:
wA ¼"�
vAvx1
w1
�2
þ�vAvx2
w2
�2
þ/þ�vAvxn
wn
�2#1=2
(1)
where A represents the calculated thermodynamic quantity,
wA is the uncertainty in the calculated result andwn stands for
a measurement accuracy or an uncertainty in one of the
dependent variables. All instruments were properly calibrated
before the experiments in order to limit systematic experi-
mental errors, while the accuracies reported within this sec-
tion for the temperature, pressure and mass flow
measurements were directly taken from manufacturer cali-
bration data. Mass balance closures corresponding to the
maximum exergy efficiency operational regimes are provided
in Table 2 in order to validate the experimental data used in
the exergy analysis. The observed closure errors are compat-
ible with the instruments accuracies and in agreement to the
ones reported by Yildiz and co-workers using a similar
experimental setup [46].
Reaction steps for the glycerol steam reforming
According to Dupont and co-authors [47], the complex re-
actions involved in the steam reforming of glycerol can be
summarised by the simplified mechanism described as fol-
lows. The first step is the decomposition of glycerol into
syngas:
C3H8O30H2O
3COþ 4H2 (2)
that is followed by the water-gas shifting reaction:
COþH2O #lowT
highTCO2 þH2 (3)
Combining (2) with 3x (3) yields the overall steam reform-
ing of glycerol:
C3H8O3 þ 3H2O03CO2 þ 7H2 (4)
that limits the hydrogen syngas concentration to 70% in
conventional steam reforming. Hydrogen can be consumed at
low temperature via methanation of CO and CO2, whereas at
high temperatures the reverse reactions of methane steam
reforming prevail [39]:
COþ 3H2 #lowT
highTCH4 þH2O (5)
Table 2 e Mass balance closures.
Feed Ratio T [�C] Input Stream [g]
Glycerol/Water Dry Syng
50% 700 150.3 60.6
70% 750 158.9 88.7
CO2 þ 4H2 #lowT
highTCH4 þ 2H2O (6)
Exergy analysis
Exergy is the amount of work obtained when a piece of matter
is brought to a state of thermodynamic equilibrium with the
common components of its surroundings by means of
reversible processes. This is a broad definition of exergy, since
thermodynamic equilibrium includes not only pressure and
temperature but also chemical equilibrium with the sub-
stances of the environment [48]. The study of complex and
chained processes can be rationalised with the aid of the
exergy concept through the so-called exergy analysis, a ther-
modynamic framework that allows the evaluation of useful
effects, resource destruction and losses to the environment on
a common basis: the amount of available work.
The molar specific exergy associated to a given flow of
matter, known as its flow exergy (e), is composed of two
distinct contributions: the thermomechanical (etm) and the
chemical parcels (ech):
e ¼ e
tm þ e
ch (7)
The thermomechanical parcel corresponds to the
maximumamount ofwork that can be obtained froma stream
when it is brought to thermal and mechanical equilibrium
with the reference environment, being given by:
etm ¼ ðh
� h
0Þ � T0ðs � s
0Þ (8)
where the specific molar properties h
0 and s0 are calculated at
the reference environment pressure (P0) and temperature (T0)
but maintaining the original stream composition unchanged
due to the exclusive thermal and mechanical equilibrium re-
striction. Under such conditions, the stream is said to be at the
restricted dead state.
In order to reach the complete equilibrium between the
stream and the ambient, further chemical iterations are
necessary to match the values of the chemical potentials of
each substance in the stream to the ones corresponding to the
presence of this same substance in the reference environ-
ment. When this condition is attained, the stream is said to
have achieved the dead state, which is sometimes referred as
unrestricted dead state. The maximum amount of work that
can be obtained as the stream goes from its restricted dead
state to the complete equilibrium with the reference envi-
ronment, reaching the dead state, corresponds to its chemical
flow exergy. The specific flow chemical exergy can be calcu-
lated from:
Output Streams [g]
as Water Tar Char Closure Error
72.5 15.2 0.8 1.2
47.8 19.3 1.1 2.0
Fig. 2 e Control volumes adopted for the thermodynamic analysis.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 79000
ech ¼
Xi
xie0;i þ h
mix þ RT0
Xi
xiln gi$xi (9)
In this equation xi represents the mole fraction of the
mixture component i, e0;i its standard chemical exergy and gi
its activity coefficient. In circumstances where the ideal gas
model is a suitable simplification, the enthalpy ofmixing (h
mix)
becomes null and the activity coefficients are unitary, what
causes Eq. (9) to be simplified to:
ech ¼
Xi
xie0;i þ R
T0
Xi
xiln $xi (10)
Two distinct control volumes were adopted for thermo-
dynamic analysis purposes, one corresponding to the heat-
recovery reforming reactor and the other to the liquid sepa-
rator, as can be seen in Fig. 2. Steady state conditions were
considered and, for the sake of clarity, the solid carbon formed
within the reforming process was assumed to be dispersed
into the non-volatile fraction of the products (condensate).
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 7 9001
The effects of feed-pumping work were neglected and the
control volume surfaces were placed at regions with envi-
ronment temperature. Furthermore, acetaldehyde was cho-
sen as a surrogate material in the modelling of the tar
properties that were demanded in exergy balances containing
wet syngas or condensate products. This simplification may
be justified by the fact that acetaldehyde has been frequently
reported as the main intermediate compound and liquid by-
product of the glycerol reforming process [49e51]. Szargut’s
reference environmentmodel and standard chemical exergies
[40] were adopted in the present work.
The irreversibility or exergy destruction rates in the
reforming reactor (_IR) can be obtained from the exergy balance
equation applied to the control volume identified in Fig. 2:
_Nfeedefeed þ _Nexin e
exin � _Nexout e
exout � _Nwsyne
wsyn � _IR ¼ 0 (11)
with the molar flow rates ( _N) and exergies (e) being calcu-
lated from experimental data and Eqs. (7) to (10). The water-
glycerin (feed) was described as a non-ideal solution with
mixing enthalpy and activity coefficients taken from Refs.
[52,53], respectively. The ideal gas model was adopted for the
engine exhaust hot gases points exin and exout, as well as for
the syngas flows (wsyn and dsyn). The thermodynamic prop-
erties of the engine exhaust gas were determined by applying
a modified version of the PER and EQMD routines proposed by
Olikara and Borman [54], as described in Ref. [55]. The wet
syngas flow was also considered as an ideal gas mixture
composed by hydrogen, carbon monoxide, carbon dioxide,
methane, ethylene, acetaldehyde, water and dispersed solid
carbon. The specific properties of the syngas constituents
were taken from Chemkin II routines [56].
The exergy balance for the liquid separator, or condenser,
can be written as follows to determine the exergy destruction:
_Nwsynewsyn � _Ndsyne
dsyn � _Nconde
cond � _IC ¼ 0 (12)
Fig. 3 e Glycerol conversion to gas phase products obtained
for the heat-recovery steam reforming of glycerol.
The condensate stream was modelled as a liquid mixture
composed by water, the tar surrogate acetaldehyde and
dispersed solid carbon. Activity coefficients were taken from
Ref. [57] and mixing enthalpies from Ref. [52]. Acetaldehyde
concentration was calculated from carbon and hydrogen
balances that were performed to the liquid separator with
data provided by gas chromatography and total organic car-
bon analysis of the filtered condensate phase.
Performance metrics
Work and heat transfer terms were not present within the
exergy balances, which were restricted to flow and irre-
versibility terms. Under these conditions, the terms of the
exergy balance equations containing exergy inflows are
regarded as exergy inputs, while the outflows can be inter-
preted as useful products or as exergy loses according to its
characteristics [40]. Accordingly, the following expression
can be written for the reforming reactor efficiency by
manipulating Eq. (11):
hR ¼_Nwsyne
wsyn
_Nfeedefeed þ _Nexin e
exin
¼ 1� DR � Lexout (13)
where the normalised values of the exergy destruction within
the reactor (DR) and exhaust gases losses (Lexout ) are as follows:
DR ¼_IR
_Nfeedefeed þ _Nexin e
exin
(14)
Lexout ¼_Nexout e
exout
_Nfeedefeed þ _Nexin e
exin
(15)
In a similar way, the liquid separator efficiency is given by:
hC ¼_Ndsyne
dsyn
_Nwsynewsyn
¼ 1� DC � Lcond (16)
with the normalised values of exergy destruction (DC) and
combined char and tar losses (Lcond) being:
DC ¼_IC
_Nwsynewsyn
(17)
Lcond ¼_Nconde
cond
_Nwsynewsyn
(18)
Finally, a global exergy efficiency expression that involves
the exergy balance for the complete reforming system can be
written by combining Eqs. (11) and (12):
hG ¼ hRhC ¼_Ndsyne
dsyn
_Nfeedefeed þ _Nexin e
exin
¼ 1� DR � Lexout � DC � Lcond (19)
The terms appearing in Eq. (19) were combined to construct
Sankey-Grassmanndiagrams,which proved to be a useful tool
for understanding the effects of operational parameters on
the performance of the reforming system.
Fig. 4 e Hydrogen and methane concentrations obtained in the dry syngas.
Fig. 5 e Syngas exergy values obtained for the heat-
recovery steam reforming of glycerol.
Fig. 6 e Exergy efficiencies obtained for the r
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 79002
Results and discussion
Synthesis gas composition and exergy content
Glycerol conversion to gas phase products, Fig. 3, reached
84.2 ± 6.4% with the reactor operating at the usual glycerol
feed concentrations levels that ranged from 10 up to 70%.
Those results corresponded well with data reported by Boba-
dilla et al. under similar bed material and operational condi-
tions [31]. Further increasing the glycerol ratio to 90% caused
water to become a limiting reactant in the wet reforming re-
actions, while the parallel pyrolysis reactions were still
restricted by the relatively low temperature conditions, thus
limiting the reforming advance and reducing the gas phase
conversion levels to about 30%. In fact, this singular opera-
tional condition was inserted in the present work in order to
better explore the energy efficiency penalties caused by low
levels of gas phase conversion, namely the low value of syngas
exergy and the higher than usual tar associated losses.
Contour maps representing the syngas compositions ob-
tained by the experiments are shown in Fig. 4. Hydrogen and
eforming reactor (hR) and condenser (hC).
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 7 9003
methane are themost relevant syngas constituents for energy
production and, from this figure, it is observed that their yields
were oppositely affected by temperature and feed concen-
tration. Methane production was enhanced at relatively low
temperatures and glycerol feed concentrations, reaching its
maximum at 30% feed concentration and 650 �C. Such con-
ditions lead to methane formation via methanation of CO and
CO2 (see Eqs. (5) and (6)), an exothermic mechanism that de-
mands hydrogen consumption. On the other hand, hydrogen
production was intensified from 50 to 70% of glycerol feed
ratios and reactor temperatures above 750 �C. At such ther-
modynamic states the exothermic methanation reactions
were inhibited in favour of the endothermic glycerol steam
reforming overall reaction (Eq. (4)). Further discussion on the
role of temperature and feed ratio on the hydrogen and
methane production through glycerol reforming can be found
in the work of Hajjaji et al. [39].
Synthesis gas with 50e54% hydrogen mole fraction and
restricted methane and carbon monoxide contents was ob-
tained at reforming temperatures from 750 to 800 �C and
glycerol feed ratios of 50e70%. When compared to the
hydrogenmole fractions of 57 ± 3% that are usually reported to
micro-scale homogeneously heated reforming reactors oper-
ating with alumina supported bedmaterial [29,30,35,50], those
results indicate a slight reduction of the hydrogen yield ob-
tained with the heat-recovery reactor configuration. This
disparity arises from the non-homogeneous bed temperature
profile that is characteristic of the counterflow heating adop-
ted here, which causes the reactor to operate with reduced
temperatures at its inlet, a condition that is susceptible to the
methane formation in detriment of hydrogen yield.
It can be observed from Fig. 5 that the exergy value of the
syngas approaches a maximum threshold of 25 ± 2 MJ/kg for
reactor outlet temperature conditions ranging from 700 to
800 �C and glycerol to water feed ratios of 50 and 70%. Taking
combustion applications into account, it is interesting to
notice that this value represents 135% of the original glycerol
exergy value. Pashchenko and co-workers defined a heat
transformation coefficient as the ratio between the lower heat
Fig. 7 e Global exergy efficiencies obtained for the heat-
recovery steam reforming of glycerol.
values of the syngas and glycerol, reporting a value of 108% at
50% glycerol water feed ratio and 627.13 �C [37]. For the sake of
comparison, our experiments indicate a heat transformation
coefficient of 98% at this same condition, a value that is 10%
lower to the one predicted by the previous authors from
equilibrium Gibbs free energy minimization simulations.
Exergy efficiencies
The reforming reactor exergy efficiency generally increased
with both glycerol to water feed ratio and reactor outlet tem-
perature, as can be noticed from Fig. 6a. These two factors
have the common effect of increasing the glycerol mass flow
rate, since higher bed temperatures also resulted in higher
injection rates to maintain the same residence time with
reduced values of syngas density. Two major sources of irre-
versibilities within the reactor were hypothesised in order to
explain the role of the glycerol inflow rate upon reactor effi-
ciency. The first would be the temperature difference between
reactants and hot exhaust gases, while the second would be
the chemical irreversibility inherent to the reforming
reactions.
Higher glycerol concentrations lead to an increase in the
temperatures of the vaporisation and inlet zones of the
reactor bed and, at the same time, to a reduction in the hot
exhaust gases temperatures by demanding a higher amount
of energy to sustain the endothermic reforming reactions in
comparison to water heating and vaporising. Both effects
contribute to a better matching of the temperature of the heat
source (combustion product gases) and the reforming tem-
perature, which reduces the amount of exergy destroyed in
the reforming/heat transfer processes as pointed out by
Simpson and Lutz [58]. Furthermore, the higher amount of
energy demanded by the reaction bed will also reduce the
exergy loss at the exhaust gases outflow as the glycerol feed
ratio increases.
On the other hand, the dilution effect of water vapour re-
duces the irreversibility rate of the reforming reactions and,
thus, the parcel of the irreversibility inherent to the reforming
reactions is expected to increase with the glycerol to water
feed ratio. An interesting insight on the magnitude of the
vaporising, heating and chemical reaction irreversibility ef-
fects can be found in the work dedicated to the ethanol
reforming by Casas-Led�on et al. [59]. These authors simulated
the reforming process by dividing it in individual steps
occurring at four separated systems: reactants mixing,
vaporising, heating and chemical reforming. Their results
indicate that increasing the ethanol feed concentration from
25 to 50% by mass would lead to a minor reduction in the
irreversibility parcel associated to the chemical reactions,
which represents only 7% of the irreversibility increase
experienced by the vaporising and heating parcels. Our
experimental results suggest that from 10 to 70% glycerol feed
ratio the heating and vaporisation effects also predominated
and the reactor efficiency (hR) was increased with the glycerol
feed ratio. However, at 90% glycerol feed concentration the
reforming reactions failed to advance and the methanation
mechanism predominated, as can be seen in Fig. 4, what
reduced the exergy extraction from the hot exhaust gases and
Fig. 8 e Sankey-Grassmann diagrams corresponding to the reactor outlet temperatures Twsyn of maximum exergy efficiency
in each feed concentration.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 79004
also caused high chemical irreversibilities that penalised the
exergy efficiency of the reforming process.
As can be seen from Fig. 6b, the condenser exergy effi-
ciency did not perform well for extreme values of feed ratio.
Low efficiency values were registered for 10% glycerol due to
the irreversibility associated to the condensation of a high
amount of water, while for the 90% glycerol regime the tar
losses due to lower reforming reaction progress imposed an
appreciable efficiency penalty.
The global reforming efficiency data shown in Fig. 7 reflects
the combination of the aforementioned factors, with the re-
gimes corresponding to 50 and 70% glycerol feed ratios
achieving the higher global exergy efficiency levels. A plateau
of about 75% of global exergy efficiency was observed from
650 �C of reactor outlet temperature, a value that proved suf-
ficient to promote the reforming reactions and restrain
methanation effects. These results agree with the maximum
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 7 9005
exergy efficiency of 78.23% reported by Dilmac and Ozkan for
industrial biogas steam reforming [60].
With regards to glycerol reforming, available literature
data indicate a maximum exergy efficiency of 66.1% at 827 �Cand 54% of glycerol feed ratio [39]. It is worth to mention that
this exergy efficiency was based on equilibrium calculations
and defined exclusively based on the hydrogen contribution to
the exergy value of the syngas. At this same hydrogen exergy
basis, the maximum exergy efficiency obtained in the present
study would be reduced to 26,0% being achieved at 70% glyc-
erol feed ratio and 800 �C. In spite of the discordance against
equilibrium predictions, our results agreed reasonably well
with experimental data reported to ethanol steam reforming
that indicate a maximum hydrogen based exergy efficiency of
24.0% [44].
Exergy balances
Further insight regarding the overall system performance can
be obtained from exergy balances applied to the reforming
reactor and liquid separator. In the interests of simplicity and
clarity, the description that follows is limited to the outlet
reactor temperatures thatmaximised hG for each glycerol feed
concentration, whereas the exergy balances are represented
by the Sankey-Grassmann diagrams depicted in Fig. 8. As it
can be noticed from this figure, the reforming reactor receives
two exergy inputs corresponding to the hot gases and feed
streams, providing the wet-syngas flow as its useful output.
Two thermodynamics imperfections are present within the
reforming reactor: The internal irreversibilities and an exergy
loss associated to the exhaust gas that is discharged to the
surroundings. In the condenser or liquid separator, the exergy
input associated to the wet syngas is derived in three terms: A
useful dry syngas output, a term of exergy destruction due to
the irreversibilities associated to the cool-down of the wet
syngas to ambient temperature (DC), and a term correspond-
ing to the exergy lost due to non-volatile tar condensates and
solid char (Lcond).
As discussed before, the glycerol content within the re-
actants improved the heat recovery from the engine exhaust
gases and, for that reason, simultaneously reduced the exergy
destruction (irreversibilities) within the reforming reactor and
the hot exhaust gases exergy losses. Thus, the amount of
exergy loss through the exhaust gases outlet ranged from 13 to
3% of the total exergy input, while the exergy destruction
ranged from 51.6 to 18.3% as the glycerol feed concentration
was increased from 10 to 90% in water. On the other hand,
water content within the reaction bed has shown to play an
important role in the reforming reaction progress, what
imposed an operational limit to the glycerol feed concentra-
tion levels. When translated into an exergy balance perspec-
tive, the lower levels of reforming reaction completion
observed with higher glycerol feed concentrations intensified
the exergy losses associated to tar and char formation. From
10 to 70% glycerol in water the condensate tar and char losses
increased from 0.7 to 7.8% of the total exergy input, while for
90% glycerol such losses corresponded to 26.1% of the exergy
input. A trade-off between the exergy destruction within the
system due to irreversibilities and the exergy losses associ-
ated to the concentrations of tar and char in the output
products was evidenced and, from the obtained data, 50%
glycerol in water has proven to be the best operational con-
dition reaching high exergy efficiency values from tempera-
tures above 700 �C.Still regarding Fig. 8, it is possible to notice that the hot
gases exergy input was reduced from 69.3 to 16% of the total
exergy input as the feed water content was diminished from
90 to 10% wt, while in the most efficient regimes the hot gases
exergy input was responsible for a quarter of the total exergy
demanded by the system. From these results it can be spec-
ulated that the efficiency benefits of the heat-recovery strat-
egy more than compensate the slight reduction experienced
in hydrogen yield due to the non-homogeneous reactor tem-
perature profile. Considering the most efficient operational
conditions, it is also interesting to point out that only 5.4e7.0%
of the total exergy input was destroyed in the condenser to
cool-down the wet syngas and extract the liquid fraction, thus
limiting the margin for efficiency gains by recovering some of
the exergy destroyed in the condenser to pre-heat the
reforming water inlet stream.
From the previous discussion, it is evident that the
reforming temperature should guarantee a suitable degree of
reaction advance in order to provide an adequate matching
between the stream temperatures at the reforming reactor
and to avoid excessive tar and char formation, while simul-
taneously maintaining a tolerable value to limit the irrevers-
ibility in the liquid separator. From Fig. 8c and d it can be
noticed that the reactor outlet temperatures corresponding to
maximum exergy efficiencywere 700 and 750 �C at 50 and 70%
glycerol feed concentrations, respectively. Conversely, for the
remaining of the studied feed concentrations lower levels of
exergy efficiency were obtained even at 800 �C. These results
suggest that the use of catalysts or other means to boost the
reforming reaction advance at lower temperatures would lead
to simultaneous reductions in the major imperfections of the
studied system: the irreversibilities within the reactor and the
condenser and the exergy losses associated to the tar and hot
exhaust gases.
Conclusion
The glycerol steam reforming was studied by means of small-
scale experiments and its degree of thermodynamic perfec-
tion was analysed through the exergy analysis method. The
experiments evidenced the preponderant role of the reactants
feed concentration with 50%e70% glycerol by weight
achieving a coincident maximum exergy efficiency value of
75.8%. As for the syngas exergy content, a maximum value of
24.3 MJ/kg was registered at 50% glycerol feed concentration
and 800 �C. This value represents 136.2% of the original glyc-
erol exergy content, constituting an important gain in com-
bustion applications. About a quarter of the total exergy input
demanded by the reforming system was recovered from hot
exhaust gases, what proved to be an interesting option with
regards to energy conservation. At the same time, the bed
temperature profile resulting from this option provided a
slightly inferior hydrogen yield in comparison to data reported
to uniformly heated reactors.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 79006
The extent of the considered thermodynamic imperfec-
tions, namely the reactor irreversibility, condenser irrevers-
ibility, exhaust gases losses and tar/char losses was very
similar at maximum efficiency conditions, ranging from 4.2 to
7.8% of the total exergy demanded for the system operation.
The importance of reactor improvements in order to reach
higher levels of reforming reaction advance/hydrogen yield
was confirmed by the exergy analysis, which concluded that
such improvements would promote a simultaneous reduction
of the main components irreversibilities and of the exergy
losses associated to exhaust gases and tar/char formation.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Acknowledgements
This study is supported by the National Council for Scientific
and Technological Development of Brazil (No. 574640).
r e f e r e n c e s
[1] I. E. Agency. World energy outlook 2019. 2019.[2] Musa SD, Zhonghua T, Ibrahim AO, Habib M. China’s energy
status: a critical look at fossils and renewable options. RenewSustain Energy Rev 2018;81:2281e90.
[3] Agarwal AK. Biofuels (alcohols and biodiesel) applications asfuels for internal combustion engines. Prog Energy CombustSci 2007;33(3):233e71.
[4] Veljkovi�c VB, Biberd�zi�c MO, Bankovi�c-Ili�c IB, Djalovi�c IG,Tasi�c MB, Nje�zi�c ZB, Stamenkovi�c OS. Biodiesel productionfrom corn oil: a review. Renew Sustain Energy Rev2018;91:531e48.
[5] Sandouqa A, Al-Hamamre Z. Energy analysis of biodieselproduction from jojoba seed oil. Renew Energy2019;130:831e42.
[6] Hajjari M, Tabatabaei M, Aghbashlo M, Ghanavati H. Areview on the prospects of sustainable biodiesel production:a global scenario with an emphasis on waste-oil biodieselutilization. Renew Sustain Energy Rev 2017;72:445e64.
[7] Bankovi�c-Ili�c IB, Stojkovi�c IJ, Stamenkovi�c OS, Veljkovic VB,Hung Y-T. Waste animal fats as feedstocks for biodieselproduction. Renew Sustain Energy Rev 2014;32:238e54.
[8] Chen J, Li J, Dong W, Zhang X, Tyagi RD, Drogui P,Surampalli RY. The potential of microalgae in biodieselproduction. Renew Sustain Energy Rev 2018;90:336e46.
[9] Rathore V, Newalkar BL, Badoni R. Processing of vegetable oilfor biofuel production through conventional and non-conventional routes. Energy for Sustainable Development2016;31:24e49.
[10] Crudo D, Bosco V, Cavagli�a G, Grillo G, Mantegna S,Cravotto G. Biodiesel production process intensificationusing a rotor-stator type generator of hydrodynamiccavitation. Ultrason Sonochem 2016;33:220e5.
[11] Kumar G, Saratale RG, Kadier A, Sivagurunathan P, Zhen G,Kim S-H, Saratale GD. A review on bio-electrochemical
systems (bess) for the syngas and value added biochemicalsproduction. Chemosphere 2017;177:84e92.
[12] Ambat I, Srivastava V, Sillanp€a€a M. Recent advancement inbiodiesel production methodologies using various feedstock:a review. Renew Sustain Energy Rev 2018;90:356e69.
[13] Escapa A, Manuel MF, Mor�an A, G�omez X, Guiot SR,Tartakovsky B. Hydrogen production from glycerol in amembraneless microbial electrolysis cell. Energy Fuels2009;23(9):4612e8.
[14] Schwengber CA, Alves HJ, Schaffner RA, da Silva FA,Sequinel R, Bach VR, Ferracin RJ. Overview of glycerolreforming for hydrogen production. Renew Sustain EnergyRev 2016;58:259e66.
[15] Anuar MR, Abdullah AZ. Challenges in biodiesel industrywith regards to feedstock, environmental, social andsustainability issues: a critical review. Renew Sustain EnergyRev 2016;58:208e23.
[16] Cornejo A, Barrio I, Campoy M, L�azaro J, Navarrete B.Oxygenated fuel additives from glycerol valorization. mainproduction pathways and effects on fuel properties andengine performance: a critical review. Renew Sustain EnergyRev 2017;79:1400e13.
[17] Monteiro MR, Kugelmeier CL, Pinheiro RS, Batalha MO, daSilva C�esar A. Glycerol from biodiesel production:Technological paths for sustainability. Renew Sustain EnergyRev 2018;88:109e22.
[18] Saladini F, Patrizi N, Pulselli FM, Marchettini N, Bastianoni S.Guidelines for emergy evaluation of first, second and thirdgeneration biofuels. Renew Sustain Energy Rev2016;66:221e7.
[19] He QS, McNutt J, Yang J. Utilization of the residual glycerolfrom biodiesel production for renewable energy generation.Renew Sustain Energy Rev 2017;71:63e76.
[20] Chozhavendhan S, Devi GK, Bharathiraja B, Kumar RP,Elavazhagan S. 9 - assessment of crude glycerol utilizationfor sustainable development of biorefineries. In: Kumar RP,Gnansounou E, Raman JK, Baskar G, editors. Refiningbiomass residues for sustainable energy and bioproducts.Academic Press; 2020. p. 195e212.
[21] Pompeo F, Santori G, Nichio NN. Hydrogen and/or syngasfrom steam reforming of glycerol. study of platinumcatalysts. Int J Hydrogen Energy 2010;35(17):8912e20.
[22] Bueno AV, de Oliveira MLM. Glycerol steam reforming in abench scale continuous flow heat recovery reactor. Int JHydrogen Energy 2013;38(32):13991e4001.
[23] Mehrpooya M, Ghorbani B, Abedi H. Biodiesel productionintegrated with glycerol steam reforming process, solid oxidefuel cell (sofc) power plant. Energy Convers Manag2020;206:112467.
[24] Ziyai MR, Mehrpooya M, Aghbashlo M, Omid M, Alsagri AS,Tabatabaei M. Techno-economic comparison of threebiodiesel production scenarios enhanced by glycerolsupercritical water reforming process. Int J Hydrogen Energy2019;44(33):17845e62.
[25] Cormos A-M, Cormos C-C. Techno-economic andenvironmental performances of glycerol reforming forhydrogen and power production with low carbon dioxideemissions. Int J Hydrogen Energy 2017;42(12):7798e810.
[26] Mattson J, Langness C, Niles B, Depcik C. Usage of glycerin-derived, hydrogen-rich syngas augmented by soybeanbiodiesel to power a biodiesel production facility. Int JHydrogen Energy 2016;41(38):17132e44.
[27] Silva JM, Soria M, Madeira LM. Challenges and strategies foroptimization of glycerol steam reforming process. RenewSustain Energy Rev 2015;42:1187e213.
[28] Cortright RD, Davda RR, Dumesic JA. Hydrogen from catalyticreforming of biomass-derived hydrocarbons in liquid water.Nature 2002;418(6901):964e7.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 6 ( 2 0 2 1 ) 8 9 9 5e9 0 0 7 9007
[29] Chen H, Ding Y, Cong NT, Dou B, Dupont V, Ghadiri M,Williams PT. A comparative study on hydrogen productionfrom steam-glycerol reforming: thermodynamics andexperimental. Renew Energy 2011;36(2):779e88.
[30] Sad M, Duarte H, Vignatti C, Padr�o C, Apesteguıa C. Steamreforming of glycerol: hydrogen production optimization. IntJ Hydrogen Energy 2015;40(18):6097e106.
[31] Bobadilla L, �Alvarez A, Domınguez M, Romero-Sarria F,Centeno M, Montes M, Odriozola J. Influence of the shape ofni catalysts in the glycerol steam reforming. Appl Catal BEnviron 2012;123e124:379e90.
[32] Macedo MS, Soria M, Madeira LM. Glycerol steam reformingfor hydrogen production: traditional versus membranereactor. Int J Hydrogen Energy 2019;44(45):24719e32.
[33] Delparish A, Koc S, Caglayan BS, Avci AK. Oxidative steamreforming of glycerol to synthesis gas in a microchannelreactor. Catal Today 2019;323:200e8.
[34] Chen D, Wang W, Liu C. Hydrogen production throughglycerol steam reforming over beehive-biomimeticgraphene-encapsulated nickel catalysts. Renew Energy2020;145:2647e57.
[35] Feng P, Huang K, Xu Q, Qi W, Xin S, Wei T, Liao L, Yan Y. Nisupported on the cao modified attapulgite as catalysts forhydrogen production from glycerol steam reforming. Int JHydrogen Energy 2020;45(15):8223e33.
[36] S. Moogi, L. Nakka, S. P. Potharaju, A. Ahmed, A. Farooq, S.-C.Jung, G. H. Rhee, Y.-K. Park, Copper promoted co/mgo: astable and efficient catalyst for glycerol steam reforming, IntJ Hydrogen Energy.
[37] Pashchenko D, Gnutikova M, Karpilov I. Comparison study ofthermochemical waste-heat recuperation by steamreforming of liquid biofuels. Int J Hydrogen Energy2020;45(7):4174e81.
[38] Hajjaji N, Baccar I, Pons M-N. Energy and exergy analysis astools for optimization of hydrogen production by glycerolautothermal reforming. Renew Energy 2014;71:368e80.
[39] Hajjaji N, Chahbani A, Khila Z, Pons M-N. A comprehensiveenergyeexergy-based assessment and parametric study of ahydrogen production process using steam glycerolreforming. Energy 2014;64:473e83.
[40] Szargut J, Morris D, Steward F. Exergy analysis of thermal,chemical, and metallurgical processes. New York, NY:Hemisphere Publishing; 1988.
[41] Chen Z, Gao L, Han W, Zhang L. Energy and exergy analysesof coal gasification with supercritical water and o2-h2o. ApplTherm Eng 2019;148:57e63.
[42] Gonz�alez WA, Zimmermann F, P�erez JF. Thermodynamicassessment of the fixed-bed downdraft gasification processof fallen leaves pelletized with glycerol as binder. CaseStudies in Thermal Engineering 2019;14:100480.
[43] Khila Z, Baccar I, Jemel I, Houas A, Hajjaji N. Energetic,exergetic and environmental life cycle assessment analysesas tools for optimization of hydrogen production byautothermal reforming of bioethanol. Int J Hydrogen Energy2016;41(39):17723e39.
[44] Hedayati A, Corre] OL, Lacarri�ere B, Llorca J. Experimentaland exergy evaluation of ethanol catalytic steam reformingin a membrane reactor. Catal Today 2016;268:68e78.
[45] Holman JP. Experimental methods for engineers. 8th ed. NewYork: McGraw-Hill; 2011.
[46] Yildiz G, Lathouwers T, Toraman HE, van Geem KM,Marin GB, Ronsse F, van Duren R, Kersten SRA, Prins W.Catalytic fast pyrolysis of pine wood: effect of successivecatalyst regeneration. Energy Fuels 2014;28(7):4560e72.
[47] Dou B, Dupont V, Rickett G, Blakeman N, Williams PT,Chen H, Ding Y, Ghadiri M. Hydrogen production bysorption-enhanced steam reforming of glycerol. BioresourTechnol 2009;100(14):3540e7.
[48] Gallo W, Milanez L. Choice of a reference state for exergeticanalysis. Energy 1990;15(2):113e21.
[49] Stein YS, Antal MJ, Jones M. A study of the gas-phasepyrolysis of glycerol. J Anal Appl Pyrol 1983;4(4):283e96.
[50] Valliyappan T, Ferdous D, Bakhshi NN, Dalai AK. Productionof hydrogen and syngas via steam gasification of glycerol in afixed-bed reactor. Top Catal 2008;49(1):59e67.
[51] Aman D, Radwan D, Ebaid M, Mikhail S, van Steen E.Comparing nickel and cobalt perovskites for steamreforming of glycerol. Molecular Catalysis 2018;452:60e7.
[52] Batov DV, Zaichikov AM, Slyusar VP, Korolev VP. Enthalpiesof mixing and state of components in aqueous-organicmixtures with nets of hydrogen bonds. Russ J Gen Chem2001;71(8):1208e14.
[53] Rudakov AM, Sergievskii VV. Activities of the components ofglycerol-water binary solutions at 298.15 k. Russ J Phys Chem2006;80(11):1804e8.
[54] Olikara C, Borman GL. A computer program for calculatingproperties of equilibrium combustion products with someapplications to i.c. engines. In: SAE technical paper. SAEInternational; 1975.
[55] Bueno AV, Vel�asquez JA, Milanez LF. Heat release and engineperformance effects of soybean oil ethyl ester blending intodiesel fuel. Energy 2011;36(6):3907e16.
[56] Kee FMRRJ, Miller JA. Chemkin-ii: a fortran chemical kineticspackage for the analysis of gas-phase chemical kinetics. In:Tech. rep. Livermore, CA (USA): Sandia National Labs.; 1989.
[57] D’Avila SG, Silva RSF. Isothermal vapor-liquid equilibriumdata by total pressure method. systems acetaldehyde-ethanol, acetaldehyde-water, and ethanol-water. J Chem EngData 1970;15(3):421e4.
[58] Simpson AP, Lutz AE. Exergy analysis of hydrogenproduction via steam methane reforming. Int J HydrogenEnergy 2007;32(18):4811e20.
[59] Casas-Led�on Y, Arteaga-Perez LE, Morales-Perez MC, Peralta-Su�arez LM. Thermodynamic analysis of the hydrogenproduction from ethanol: first and second laws approaches.ISRN Thermodynamics 2012;2012:672691.
[60] Dilmac OF, Ozkan SK. Energy and exergy analyses of a steamreforming process for hydrogen production. Int J Exergy2008;5(2):241e8.